docker: Error response from daemon: could not select device driver ““ with capabilities: [[gpu]]

在docker创建容器加 --gpus all 的时候：

docker run -itd --rm --name spvconv1 --gpus all  -v /syy:/home/lcbryant/syy royokong/torch1.6.0-cuda10.1-apex-base:latest

报出如下错误：

docker: Error response from daemon: could not select device driver "" with capabilities: [[gpu]].

解决方案

1.安装nvidia驱动

确认已安装NVIDIA卡

lspci -vv | grep -i nvidia

2.安装NVIDIA Container Runtime

先预装好一些环境，新建一个nvidia-container-runtime-script.sh，内容如下：

sudo curl -s -L https://nvidia.github.io/nvidia-container-runtime/gpgkey | \
  sudo apt-key add -
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
sudo curl -s -L https://nvidia.github.io/nvidia-container-runtime/$distribution/nvidia-container-runtime.list | \
  sudo tee /etc/apt/sources.list.d/nvidia-container-runtime.list
sudo apt-get update

然后运行：

sudo bash nvidia-container-runtime-script.sh

安装NVIDIA Container Runtime：

sudo apt-get install nvidia-container-runtime

查看是否装好：

which nvidia-container-runtime-hook

结果是：

3.验证

验证docker是否装好：

docker version

验证一下gpus选项：

docker run --help | grep -i gpus

最后再重启一下！

systemctl restart docker

然后就可以愉快地在docker容器里使用gpu啦！